selftests/sgx: Trigger the page reclaimer and #PF handler
Subsystems (4)
Jarkko Sakkinen
maintainercover selftests/sgx: Trigger the page reclaimer and #PF handler
Create a heap for the test enclave, which has the same size as all available Enclave Page Cache (EPC) pages in the system. This will guarantee that all test_encl.elf pages *and* SGX Enclave Control Structure (SECS) have been swapped out by the page reclaimer during the load time. Actually, this adds a bit more stress than that since part of the EPC gets reserved for the Version Array (VA) pages.
For each test, the page fault handler gets triggered in two occasions:
- When SGX_IOC_ENCLAVE_INIT is performed, SECS gets swapped in by the page fault handler.
- During the execution, each page that is referenced gets swapped in by the page fault handler.
Jarkko Sakkinen (3):
x86/sgx: Add sgx_nr_all_pages to the debugfs selftests/sgx: Assign source for each segment selftests/sgx: Trigger the reclaimer and #PF handler
Tianjia Zhang (1):
selftests/sgx: Fix Q1 and Q2 calculation in sigstruct.c
Documentation/x86/sgx.rst | 9 +++++ arch/x86/kernel/cpu/sgx/main.c | 10 ++++- tools/testing/selftests/sgx/load.c | 38 ++++++++++++++---- tools/testing/selftests/sgx/main.c | 42 +++++++++++++++++++- tools/testing/selftests/sgx/main.h | 4 +- tools/testing/selftests/sgx/sigstruct.c | 53 +++++++++++++------------ 6 files changed, 120 insertions(+), 36 deletions(-)
Signature · 1 line
2.32.0
Jarkko Sakkinen
maintainerPATCH 1/4 x86/sgx: Add sgx_nr_all_pages to the debugfs
Create /sys/kernel/debug/x86/sgx_nr_all_pages, which reports total number of EPC pages available in the system.
--- Documentation/x86/sgx.rst | 9 +++++++++ arch/x86/kernel/cpu/sgx/main.c | 10 +++++++++- 2 files changed, 18 insertions(+), 1 deletion(-)
Documentation/x86/sgx.rst +9 -0
@@ -250,3 +250,12 @@ user wants to deploy SGX applications both on the host and in guests on the same machine, the user should reserve enough EPC (by taking out total virtual EPC size of all SGX VMs from the physical EPC size) for host SGX applications so they can run with acceptable performance. + +SGX debugging +============= + +The total number of available EPC pages can observed by: + + # mount -t debugfs debugfs /sys/kernel/debug + # cat /sys/kernel/debug/x86/sgx_nr_all_pages + <the number of all EPC pages available in the system>
arch/x86/kernel/cpu/sgx/main.c +9 -1
@@ -1,6 +1,7 @@// SPDX-License-Identifier: GPL-2.0 /* Copyright(c) 2016-20 Intel Corporation. */ +#include <linux/debugfs.h> #include <linux/file.h> #include <linux/freezer.h> #include <linux/highmem.h>
@@ -28,7 +29,10 @@ static DECLARE_WAIT_QUEUE_HEAD(ksgxd_waitq);static LIST_HEAD (sgx_active_page_list);static DEFINE_SPINLOCK (sgx_reclaimer_lock); -/* The free page list lock protected variables prepend the lock. */ +/* The number of EPC pages in total in all nodes. */ +static unsigned long sgx_nr_all_pages; + +/* The number of free EPC pages in all nodes. */ static unsigned long sgx_nr_free_pages;/* Nodes with one or more EPC sections. */
@@ -656,6 +660,8 @@ static bool __init sgx_setup_epc_section(u64 phys_addr, u64 size,list_add_tail (§ion->pages[i].list, &sgx_dirty_page_list); } + sgx_nr_all_pages += nr_pages; +return true; }
@@ -823,6 +829,8 @@ static int __init sgx_init(void)if (sgx_vepc_init () && ret) goto err_provision; +debugfs_create_ulong ("sgx_nr_all_pages" ,0444 , arch_debugfs_dir, &sgx_nr_all_pages); +return 0 ; err_provision:
Signature · 1 line
2.32.0
Re: [PATCH 1/4] x86/sgx: Add sgx_nr_all_pages to the debugfs
Quoted 2 lines · Jarkko Sakkinen
Create
/sys/kernel/debug/x86/sgx_nr_all_pages, which reports total number of EPC pages available in the system.
Could we flesh this out a bit, please?
What's the value here when userspace could just re-enumerate the EPC size from CPUID?
I'd really appreciate if we could draw parallels between these additions to the "SGX VM" and their analogs in the "core VM". In this case, I think the closest analog is probably "MemTotal" in /proc/meminfo.
Second, how is this going to be used?
Third, is this going to be the ABI forever?
Re: [PATCH 1/4] x86/sgx: Add sgx_nr_all_pages to the debugfs
Quoted 8 lines · Dave Hansen
Quoted 2 lines · Jarkko Sakkinen
Create
/sys/kernel/debug/x86/sgx_nr_all_pages, which reports total number of EPC pages available in the system.Could we flesh this out a bit, please?
What's the value here when userspace could just re-enumerate the EPC size from CPUID?
Quoted 7 lines
I'd really appreciate if we could draw parallels between these additions to the "SGX VM" and their analogs in the "core VM". In this case, I think the closest analog is probably "MemTotal" in
/proc/meminfo.Second, how is this going to be used?
Third, is this going to be the ABI forever?
debugfs is never a stable abi. If it is being used as such, then the kernel code is wrong. This better just be debugging stuff only.
thanks,
greg k-h
Jarkko Sakkinen
maintainerRe: [PATCH 1/4] x86/sgx: Add sgx_nr_all_pages to the debugfs
In reply to Dave HansenQuoted 7 lines · Dave Hansen
Quoted 2 lines · Jarkko Sakkinen
Create
/sys/kernel/debug/x86/sgx_nr_all_pages, which reports total number of EPC pages available in the system.Could we flesh this out a bit, please?
What's the value here when userspace could just re-enumerate the EPC size from CPUID?
My thinking is that it is better to use "kernel synthesized" value for the EPC size, because kernel controls the EPC.
Quoted 3 lines
I'd really appreciate if we could draw parallels between these additions to the "SGX VM" and their analogs in the "core VM". In this case, I think the closest analog is probably "MemTotal" in
/proc/meminfo.
Would make sense.
Quoted 1 line
Second, how is this going to be used?
SGX kselftest creates a heap, of which size is the same as the total size of the EPC reported by the kernel.
Quoted 1 line
Third, is this going to be the ABI forever?
AFAIK, debugfs is not part of the ABI.
/Jarkko
Jarkko Sakkinen
maintainerPATCH 2/4 selftests/sgx: Fix Q1 and Q2 calculation in sigstruct.c
In reply to Jarkko SakkinenFrom: Tianjia Zhang <[email protected]>
Q1 and Q2 are numbers with *maximum* length of 384 bytes. If the calculated length of Q1 and Q2 is less than 384 bytes, things will go wrong.
E.g. if Q2 is 383 bytes, then
1. The bytes of q2 are copied to sigstruct->q2 in calc_q1q2().
2. The entire sigstruct->q2 is reversed, which results it being 256 * Q2, given that the last byte of sigstruct->q2 is added to before the bytes given by calc_q1q2().
Either change in key or measurement can trigger the bug. E.g. an unmeasured heap could cause a devastating change in Q1 or Q2.
Reverse exactly the bytes of Q1 and Q2 in calc_q1q2() before returning to the caller.
---
The original patch did a bad job explaining the code change but it turned out making sense. I wrote a new description.
v3:
- Adjusted the fixes tag to refer to the correct commit
v2:
- Added a fixes tag.
tools/testing/selftests/sgx/sigstruct.c | 41 +++++++++++++------------ 1 file changed, 21 insertions(+), 20 deletions(-)
tools/testing/selftests/sgx/sigstruct.c +21 -20
@@ -55,10 +55,27 @@ static bool alloc_q1q2_ctx(const uint8_t *s, const uint8_t *m,return true; } +static void reverse_bytes (void *data,int length) +{ +int i =0 ; +int j = length -1 ; +uint8_t temp; +uint8_t *ptr = data; + +while (i < j) { + temp = ptr[i]; + ptr[i] = ptr[j]; + ptr[j] = temp; + i++; + j--; + } +} +static bool calc_q1q2 (const uint8_t *s,const uint8_t *m,uint8_t *q1,uint8_t *q2) {struct q1q2_ctx ctx; +int len;if (!alloc_q1q2_ctx (s, m, &ctx)) {fprintf (stderr,"Not enough memory for Q1Q2 calculation\n" );
@@ -89,8 +106,10 @@ static bool calc_q1q2(const uint8_t *s, const uint8_t *m, uint8_t *q1, goto out; } -BN_bn2bin (ctx.q1, q1); -BN_bn2bin (ctx.q2, q2); + len =BN_bn2bin (ctx.q1, q1); +reverse_bytes (q1, len); + len =BN_bn2bin (ctx.q2, q2); +reverse_bytes (q2, len);free_q1q2_ctx (&ctx);return true;
@@ -152,22 +171,6 @@ static RSA *gen_sign_key(void)return key; } -static void reverse_bytes (void *data,int length) -{ -int i =0 ; -int j = length -1 ; -uint8_t temp; -uint8_t *ptr = data; - -while (i < j) { - temp = ptr[i]; - ptr[i] = ptr[j]; - ptr[j] = temp; - i++; - j--; - } -} -enum mrtags {MRECREATE =0x0045544145524345 ,MREADD =0x0000000044444145 ,
@@ -367,8 +370,6 @@ bool encl_measure(struct encl *encl)/* BE -> LE */ reverse_bytes (sigstruct->signature,SGX_MODULUS_SIZE );reverse_bytes (sigstruct->modulus,SGX_MODULUS_SIZE ); -reverse_bytes (sigstruct->q1,SGX_MODULUS_SIZE ); -reverse_bytes (sigstruct->q2,SGX_MODULUS_SIZE );EVP_MD_CTX_destroy (ctx);RSA_free (key);
Signature · 1 line
2.32.0
Re: [PATCH 2/4] selftests/sgx: Fix Q1 and Q2 calculation in sigstruct.c
Quoted 15 lines · Jarkko Sakkinen
From: Tianjia Zhang <[email protected]>
Q1 and Q2 are numbers with *maximum* length of 384 bytes. If the calculated length of Q1 and Q2 is less than 384 bytes, things will go wrong.
E.g. if Q2 is 383 bytes, then
1. The bytes of q2 are copied to
sigstruct->q2incalc_q1q2().
2. The entiresigstruct->q2is reversed, which results it being 256 * Q2, given that the last byte ofsigstruct->q2is added to before the bytes given bycalc_q1q2().Either change in key or measurement can trigger the bug. E.g. an unmeasured heap could cause a devastating change in Q1 or Q2.
Quoted 7 lines
Reverse exactly the bytes of Q1 and Q2 in
calc_q1q2()before returning to the caller.Fixes 2adcba79e69d ("selftests/x86: Add a selftest for SGX")Signed-off-by Tianjia Zhang <[email protected]>Signed-off-by Jarkko Sakkinen <[email protected]>
This looks fine, but can I suggest a Subject: tweak?
selftests/sgx: Fix calculations for sub-maximum field sizes
In any case:
Jarkko Sakkinen
maintainerRe: [PATCH 2/4] selftests/sgx: Fix Q1 and Q2 calculation in sigstruct.c
Quoted 19 lines · Dave Hansen
Quoted 18 lines · Jarkko Sakkinen
From: Tianjia Zhang <[email protected]>
Q1 and Q2 are numbers with *maximum* length of 384 bytes. If the calculated length of Q1 and Q2 is less than 384 bytes, things will go wrong.
E.g. if Q2 is 383 bytes, then
1. The bytes of q2 are copied to
sigstruct->q2incalc_q1q2().
2. The entiresigstruct->q2is reversed, which results it being 256 * Q2, given that the last byte ofsigstruct->q2is added to before the bytes given bycalc_q1q2().Either change in key or measurement can trigger the bug. E.g. an unmeasured heap could cause a devastating change in Q1 or Q2.
Reverse exactly the bytes of Q1 and Q2 in
calc_q1q2()before returning to the caller.
Quoted 8 lines
Quoted 4 lines
Fixes 2adcba79e69d ("selftests/x86: Add a selftest for SGX")Signed-off-by Tianjia Zhang <[email protected]>Signed-off-by Jarkko Sakkinen <[email protected]>This looks fine, but can I suggest a Subject: tweak?
selftests/sgx: Fix calculations for sub-maximum field sizes
WFM
Quoted 4 lines
In any case:
Reviewed-by Dave Hansen <[email protected]>
Thank you.
/Jarkko
Jarkko Sakkinen
maintainerPATCH 3/4 selftests/sgx: Assign source for each segment
In reply to Jarkko SakkinenDefine source per segment so that enclave pages can be added from different sources, e.g. anonymous VMA for zero pages. In other words, add 'src' field to struct encl_segment, and assign it to 'encl->src' for pages inherited from the enclave binary.
--- tools/testing/selftests/sgx/load.c | 5 +++-- tools/testing/selftests/sgx/main.h | 1 + tools/testing/selftests/sgx/sigstruct.c | 8 ++++---- 3 files changed, 8 insertions(+), 6 deletions(-)
tools/testing/selftests/sgx/load.c +3 -2
@@ -107,7 +107,7 @@ static bool encl_ioc_add_pages(struct encl *encl, struct encl_segment *seg)memset (&secinfo,0 ,sizeof (secinfo)); secinfo.flags = seg->flags; - ioc.src = (uint64_t )encl->src + seg->offset; + ioc.src = (uint64_t )seg->src; ioc.offset = seg->offset; ioc.length = seg->size; ioc.secinfo = (unsigned long )&secinfo;
@@ -226,6 +226,7 @@ bool encl_load(const char *path, struct encl *encl)if (j ==0 ) { src_offset = phdr->p_offset &PAGE_MASK ; + encl->src = encl->bin + src_offset; seg->prot =PROT_READ |PROT_WRITE ; seg->flags =SGX_PAGE_TYPE_TCS <<8 ;
@@ -238,13 +239,13 @@ bool encl_load(const char *path, struct encl *encl) seg->offset = (phdr->p_offset &PAGE_MASK ) - src_offset; seg->size = (phdr->p_filesz +PAGE_SIZE -1 ) &PAGE_MASK ; + seg->src = encl->src + seg->offset; j++; }assert (j == encl->nr_segments); - encl->src = encl->bin + src_offset; encl->src_size = encl->segment_tbl[j -1 ].offset + encl->segment_tbl[j -1 ].size;
tools/testing/selftests/sgx/main.h +1 -0
@@ -7,6 +7,7 @@#define MAIN_H struct encl_segment { +void *src;off_t offset;size_t size;unsigned int prot;
tools/testing/selftests/sgx/sigstruct.c +4 -4
@@ -289,14 +289,14 @@ static bool mrenclave_eextend(EVP_MD_CTX *ctx, uint64_t offset,static bool mrenclave_segment (EVP_MD_CTX *ctx,struct encl *encl,struct encl_segment *seg) { -uint64_t end = seg->offset + seg->size; +uint64_t end = seg->size;uint64_t offset; -for (offset = seg->offset; offset < end; offset +=PAGE_SIZE ) { -if (!mrenclave_eadd (ctx, offset, seg->flags)) +for (offset =0 ; offset < end; offset +=PAGE_SIZE ) { +if (!mrenclave_eadd (ctx, seg->offset + offset, seg->flags))return false; -if (!mrenclave_eextend (ctx, offset, encl->src + offset)) +if (!mrenclave_eextend (ctx, seg->offset + offset, seg->src + offset))return false; }
Signature · 1 line
2.32.0
Jarkko Sakkinen
maintainerPATCH 4/4 selftests/sgx: Trigger the reclaimer and #PF handler
In reply to Jarkko SakkinenCreate a heap for the test enclave, which has the same size as all available Enclave Page Cache (EPC) pages in the system. This will guarantee that all test_encl.elf pages *and* SGX Enclave Control Structure (SECS) have been swapped out by the page reclaimer during the load time. Actually, this adds a bit more stress than that since part of the EPC gets reserved for the Version Array (VA) pages.
For each test, the page fault handler gets triggered in two occasions:
- When SGX_IOC_ENCLAVE_INIT is performed, SECS gets swapped in by the page fault handler.
- During the execution, each page that is referenced gets swapped in by the page fault handler.
--- tools/testing/selftests/sgx/load.c | 33 +++++++++++++++---- tools/testing/selftests/sgx/main.c | 42 ++++++++++++++++++++++++- tools/testing/selftests/sgx/main.h | 3 +- tools/testing/selftests/sgx/sigstruct.c | 6 ++-- 4 files changed, 74 insertions(+), 10 deletions(-)
tools/testing/selftests/sgx/load.c +27 -6
@@ -21,6 +21,8 @@void encl_delete (struct encl *encl) { +struct encl_segment *heap_seg = &encl->segment_tbl[encl->nr_segments -1 ]; +if (encl->encl_base)munmap ((void *)encl->encl_base, encl->encl_size);
@@ -30,6 +32,8 @@ void encl_delete(struct encl *encl)if (encl->fd)close (encl->fd); +munmap (heap_seg->src, heap_seg->size); +if (encl->segment_tbl)free (encl->segment_tbl);
@@ -111,7 +115,10 @@ static bool encl_ioc_add_pages(struct encl *encl, struct encl_segment *seg) ioc.offset = seg->offset; ioc.length = seg->size; ioc.secinfo = (unsigned long )&secinfo; - ioc.flags =SGX_PAGE_MEASURE ; +if (seg->measure) + ioc.flags =SGX_PAGE_MEASURE ; +else + ioc.flags =0 ; rc =ioctl (encl->fd,SGX_IOC_ENCLAVE_ADD_PAGES , &ioc);if (rc <0 ) {
@@ -124,9 +131,10 @@ static bool encl_ioc_add_pages(struct encl *encl, struct encl_segment *seg) -bool encl_load (const char *path,struct encl *encl) +bool encl_load (const char *path,struct encl *encl,unsigned long heap_size) {const char device_path[] ="/dev/sgx_enclave" ; +struct encl_segment *seg;Elf64_Phdr *phdr_tbl;off_t src_offset;Elf64_Ehdr *ehdr;
@@ -188,6 +196,8 @@ bool encl_load(const char *path, struct encl *encl) ehdr = encl->bin; phdr_tbl = encl->bin + ehdr->e_phoff; + encl->nr_segments =1 ;/* one for the heap */ +for (i =0 ; i < ehdr->e_phnum; i++) {Elf64_Phdr *phdr = &phdr_tbl[i];
@@ -203,7 +213,6 @@ bool encl_load(const char *path, struct encl *encl)for (i =0 , j =0 ; i < ehdr->e_phnum; i++) {Elf64_Phdr *phdr = &phdr_tbl[i];unsigned int flags = phdr->p_flags; -struct encl_segment *seg;if (phdr->p_type !=PT_LOAD )continue ;
@@ -240,14 +249,26 @@ bool encl_load(const char *path, struct encl *encl) seg->offset = (phdr->p_offset &PAGE_MASK ) - src_offset; seg->size = (phdr->p_filesz +PAGE_SIZE -1 ) &PAGE_MASK ; seg->src = encl->src + seg->offset; + seg->measure = true; j++; } -assert (j == encl->nr_segments); +assert (j == encl->nr_segments -1 ); + + seg = &encl->segment_tbl[j]; + seg->offset = encl->segment_tbl[j -1 ].offset + encl->segment_tbl[j -1 ].size; + seg->size = heap_size; + seg->src =mmap (NULL , heap_size,PROT_READ |PROT_WRITE , +MAP_ANONYMOUS |MAP_PRIVATE ,-1 ,0 ); + seg->prot =PROT_READ |PROT_WRITE ; + seg->flags = (SGX_PAGE_TYPE_REG <<8 ) | seg->prot; + seg->measure = false; + +if (seg->src ==MAP_FAILED ) + goto err; - encl->src_size = encl->segment_tbl[j -1 ].offset + - encl->segment_tbl[j -1 ].size; + encl->src_size = encl->segment_tbl[j].offset + encl->segment_tbl[j].size;for (encl->encl_size =4096 ; encl->encl_size < encl->src_size; ) encl->encl_size <<=1 ;
tools/testing/selftests/sgx/main.c +41 -1
58 lines not on this page · raw message
tools/testing/selftests/sgx/main.h +2 -1
@@ -12,6 +12,7 @@ struct encl_segment {size_t size;unsigned int prot;unsigned int flags; +bool measure; };struct encl {
@@ -32,7 +33,7 @@ extern unsigned char sign_key[];extern unsigned char sign_key_end[];void encl_delete (struct encl *ctx); -bool encl_load (const char *path,struct encl *encl); +bool encl_load (const char *path,struct encl *encl,unsigned long heap_size);bool encl_measure (struct encl *encl);bool encl_build (struct encl *encl);
tools/testing/selftests/sgx/sigstruct.c +4 -2
@@ -296,8 +296,10 @@ static bool mrenclave_segment(EVP_MD_CTX *ctx, struct encl *encl,if (!mrenclave_eadd (ctx, seg->offset + offset, seg->flags))return false; -if (!mrenclave_eextend (ctx, seg->offset + offset, seg->src + offset)) -return false; +if (seg->measure) { +if (!mrenclave_eextend (ctx, seg->offset + offset, seg->src + offset)) +return false; + } }return true;
Signature · 1 line
2.32.0
Re: [PATCH 4/4] selftests/sgx: Trigger the reclaimer and #PF handler
Hi Jarkko,
Quoted 7 lines · Jarkko Sakkinen
Create a heap for the test enclave, which has the same size as all available Enclave Page Cache (EPC) pages in the system. This will guarantee that all test_encl.elf pages *and* SGX Enclave Control Structure (SECS) have been swapped out by the page reclaimer during the load time. Actually, this adds a bit more stress than that since part of the EPC gets reserved for the Version Array (VA) pages.
Quoted 7 lines
For each test, the page fault handler gets triggered in two occasions:
- When
SGX_IOC_ENCLAVE_INITis performed, SECS gets swapped in by the page fault handler.
- During the execution, each page that is referenced gets swapped in by the page fault handler.
If I understand this correctly, all EPC pages are now being consumed during fixture setup and thus every SGX test, no matter how big or small, now becomes a stress test of the reclaimer instead of there being a unique reclaimer test. Since an enclave is set up and torn down for every test this seems like a significant addition. It also seems like this would impact future tests of dynamic page addition where not all scenarios could be tested with all EPC pages already consumed.
Reinette
Jarkko Sakkinen
maintainerRe: [PATCH 4/4] selftests/sgx: Trigger the reclaimer and #PF handler
Quoted 26 lines · Reinette Chatre
Hi Jarkko,
Quoted 7 lines · Jarkko Sakkinen
Create a heap for the test enclave, which has the same size as all available Enclave Page Cache (EPC) pages in the system. This will guarantee that all test_encl.elf pages *and* SGX Enclave Control Structure (SECS) have been swapped out by the page reclaimer during the load time. Actually, this adds a bit more stress than that since part of the EPC gets reserved for the Version Array (VA) pages.
Quoted 7 lines
For each test, the page fault handler gets triggered in two occasions:
- When
SGX_IOC_ENCLAVE_INITis performed, SECS gets swapped in by the page fault handler.
- During the execution, each page that is referenced gets swapped in by the page fault handler.If I understand this correctly, all EPC pages are now being consumed during fixture setup and thus every SGX test, no matter how big or small, now becomes a stress test of the reclaimer instead of there being a unique reclaimer test. Since an enclave is set up and torn down for every test this seems like a significant addition. It also seems like this would impact future tests of dynamic page addition where not all scenarios could be tested with all EPC pages already consumed.
Quoted 1 line
Reinette
Re-initializing the test enclave is mandatory thing to do for all tests because it has an internals state.
/Jarkko
Re: [PATCH 4/4] selftests/sgx: Trigger the reclaimer and #PF handler
Hi Jarkko,
Quoted 27 lines · Jarkko Sakkinen
Quoted 26 lines · Reinette Chatre
Hi Jarkko,
Quoted 7 lines · Jarkko Sakkinen
Create a heap for the test enclave, which has the same size as all available Enclave Page Cache (EPC) pages in the system. This will guarantee that all test_encl.elf pages *and* SGX Enclave Control Structure (SECS) have been swapped out by the page reclaimer during the load time. Actually, this adds a bit more stress than that since part of the EPC gets reserved for the Version Array (VA) pages.
Quoted 7 lines
For each test, the page fault handler gets triggered in two occasions:
- When
SGX_IOC_ENCLAVE_INITis performed, SECS gets swapped in by the page fault handler.
- During the execution, each page that is referenced gets swapped in by the page fault handler.If I understand this correctly, all EPC pages are now being consumed during fixture setup and thus every SGX test, no matter how big or small, now becomes a stress test of the reclaimer instead of there being a unique reclaimer test. Since an enclave is set up and torn down for every test this seems like a significant addition. It also seems like this would impact future tests of dynamic page addition where not all scenarios could be tested with all EPC pages already consumed.
Quoted 5 lines
Quoted 1 line
Reinette
Re-initializing the test enclave is mandatory thing to do for all tests because it has an internals state.
Right, but not all tests require the same enclave. In kselftest terminology I think you are attempting to force all tests to depend on the same test fixture. Is it not possible to have a separate "reclaimer" test fixture that would build an enclave with a large heap and then have reclaimer tests that exercise it by being tests that are specific to this "reclaimer fixture"?
Reinette
Jarkko Sakkinen
maintainerRe: [PATCH 4/4] selftests/sgx: Trigger the reclaimer and #PF handler
Quoted 36 lines · Reinette Chatre
Hi Jarkko,
Quoted 27 lines · Jarkko Sakkinen
Quoted 26 lines · Reinette Chatre
Hi Jarkko,
Quoted 7 lines · Jarkko Sakkinen
Create a heap for the test enclave, which has the same size as all available Enclave Page Cache (EPC) pages in the system. This will guarantee that all test_encl.elf pages *and* SGX Enclave Control Structure (SECS) have been swapped out by the page reclaimer during the load time. Actually, this adds a bit more stress than that since part of the EPC gets reserved for the Version Array (VA) pages.
Quoted 7 lines
For each test, the page fault handler gets triggered in two occasions:
- When
SGX_IOC_ENCLAVE_INITis performed, SECS gets swapped in by the page fault handler.
- During the execution, each page that is referenced gets swapped in by the page fault handler.If I understand this correctly, all EPC pages are now being consumed during fixture setup and thus every SGX test, no matter how big or small, now becomes a stress test of the reclaimer instead of there being a unique reclaimer test. Since an enclave is set up and torn down for every test this seems like a significant addition. It also seems like this would impact future tests of dynamic page addition where not all scenarios could be tested with all EPC pages already consumed.
Quoted 5 lines
Quoted 1 line
Reinette
Re-initializing the test enclave is mandatory thing to do for all tests because it has an internals state.
Quoted 7 lines
Right, but not all tests require the same enclave. In kselftest terminology I think you are attempting to force all tests to depend on the same test fixture. Is it not possible to have a separate "reclaimer" test fixture that would build an enclave with a large heap and then have reclaimer tests that exercise it by being tests that are specific to this "reclaimer fixture"?
Reinette
Why add that complexity?
/Jarkko
Re: [PATCH 4/4] selftests/sgx: Trigger the reclaimer and #PF handler
Hi Jarkko,
Quoted 43 lines · Jarkko Sakkinen
Quoted 42 lines · Reinette Chatre
Hi Jarkko,
Quoted 27 lines · Jarkko Sakkinen
Quoted 26 lines · Reinette Chatre
Hi Jarkko,
Quoted 7 lines · Jarkko Sakkinen
Create a heap for the test enclave, which has the same size as all available Enclave Page Cache (EPC) pages in the system. This will guarantee that all test_encl.elf pages *and* SGX Enclave Control Structure (SECS) have been swapped out by the page reclaimer during the load time. Actually, this adds a bit more stress than that since part of the EPC gets reserved for the Version Array (VA) pages.
Quoted 7 lines
For each test, the page fault handler gets triggered in two occasions:
- When
SGX_IOC_ENCLAVE_INITis performed, SECS gets swapped in by the page fault handler.
- During the execution, each page that is referenced gets swapped in by the page fault handler.If I understand this correctly, all EPC pages are now being consumed during fixture setup and thus every SGX test, no matter how big or small, now becomes a stress test of the reclaimer instead of there being a unique reclaimer test. Since an enclave is set up and torn down for every test this seems like a significant addition. It also seems like this would impact future tests of dynamic page addition where not all scenarios could be tested with all EPC pages already consumed.
Quoted 5 lines
Quoted 1 line
Reinette
Re-initializing the test enclave is mandatory thing to do for all tests because it has an internals state.
Right, but not all tests require the same enclave. In kselftest terminology I think you are attempting to force all tests to depend on the same test fixture. Is it not possible to have a separate "reclaimer" test fixture that would build an enclave with a large heap and then have reclaimer tests that exercise it by being tests that are specific to this "reclaimer fixture"?
Quoted 4 lines
Quoted 1 line
Reinette
Why add that complexity?
With this change every test is turned into a pseudo reclaimer test without there being any explicit testing (with pass/fail criteria) of reclaimer behavior. This is an expensive addition and reduces the scenarios that the tests can exercise.
Reinette
Jarkko Sakkinen
maintainerRe: [PATCH 4/4] selftests/sgx: Trigger the reclaimer and #PF handler
Quoted 51 lines · Reinette Chatre
Hi Jarkko,
Quoted 43 lines · Jarkko Sakkinen
Quoted 42 lines · Reinette Chatre
Hi Jarkko,
Quoted 27 lines · Jarkko Sakkinen
Quoted 26 lines · Reinette Chatre
Hi Jarkko,
Quoted 7 lines · Jarkko Sakkinen
Create a heap for the test enclave, which has the same size as all available Enclave Page Cache (EPC) pages in the system. This will guarantee that all test_encl.elf pages *and* SGX Enclave Control Structure (SECS) have been swapped out by the page reclaimer during the load time. Actually, this adds a bit more stress than that since part of the EPC gets reserved for the Version Array (VA) pages.
Quoted 7 lines
For each test, the page fault handler gets triggered in two occasions:
- When
SGX_IOC_ENCLAVE_INITis performed, SECS gets swapped in by the page fault handler.
- During the execution, each page that is referenced gets swapped in by the page fault handler.If I understand this correctly, all EPC pages are now being consumed during fixture setup and thus every SGX test, no matter how big or small, now becomes a stress test of the reclaimer instead of there being a unique reclaimer test. Since an enclave is set up and torn down for every test this seems like a significant addition. It also seems like this would impact future tests of dynamic page addition where not all scenarios could be tested with all EPC pages already consumed.
Quoted 5 lines
Quoted 1 line
Reinette
Re-initializing the test enclave is mandatory thing to do for all tests because it has an internals state.
Right, but not all tests require the same enclave. In kselftest terminology I think you are attempting to force all tests to depend on the same test fixture. Is it not possible to have a separate "reclaimer" test fixture that would build an enclave with a large heap and then have reclaimer tests that exercise it by being tests that are specific to this "reclaimer fixture"?
Quoted 4 lines
Quoted 1 line
Reinette
Why add that complexity?
Quoted 6 lines
With this change every test is turned into a pseudo reclaimer test without there being any explicit testing (with pass/fail criteria) of reclaimer behavior. This is an expensive addition and reduces the scenarios that the tests can exercise.
Reinette
There is consistent known behaviour how reclaimer and also the page fault are exercised for each test. I think that is what matters most right now that the basic behaviour of both the page reclaimer and page fault handler gets exercised.
I don't understand the real-world gain of doing something factors more complex than necessary at a particular point of time, when you don't really need to hang yourself into it forever.
This patch does increase the coverage in a deterministic manner to the code paths that were not previously exercised, i.e. we know the code paths, and could even calculate the exact number of times that they are triggered. And without doing anything obscure. That's what matters to me.
/Jarkko
/Jarkko
Re: [PATCH 4/4] selftests/sgx: Trigger the reclaimer and #PF handler
Hi Jarkko,
Quoted 57 lines · Jarkko Sakkinen
Quoted 56 lines · Reinette Chatre
Hi Jarkko,
Quoted 43 lines · Jarkko Sakkinen
Quoted 42 lines · Reinette Chatre
Hi Jarkko,
Quoted 27 lines · Jarkko Sakkinen
Quoted 26 lines · Reinette Chatre
Hi Jarkko,
Quoted 7 lines · Jarkko Sakkinen
Create a heap for the test enclave, which has the same size as all available Enclave Page Cache (EPC) pages in the system. This will guarantee that all test_encl.elf pages *and* SGX Enclave Control Structure (SECS) have been swapped out by the page reclaimer during the load time. Actually, this adds a bit more stress than that since part of the EPC gets reserved for the Version Array (VA) pages.
Quoted 7 lines
For each test, the page fault handler gets triggered in two occasions:
- When
SGX_IOC_ENCLAVE_INITis performed, SECS gets swapped in by the page fault handler.
- During the execution, each page that is referenced gets swapped in by the page fault handler.If I understand this correctly, all EPC pages are now being consumed during fixture setup and thus every SGX test, no matter how big or small, now becomes a stress test of the reclaimer instead of there being a unique reclaimer test. Since an enclave is set up and torn down for every test this seems like a significant addition. It also seems like this would impact future tests of dynamic page addition where not all scenarios could be tested with all EPC pages already consumed.
Quoted 5 lines
Quoted 1 line
Reinette
Re-initializing the test enclave is mandatory thing to do for all tests because it has an internals state.
Right, but not all tests require the same enclave. In kselftest terminology I think you are attempting to force all tests to depend on the same test fixture. Is it not possible to have a separate "reclaimer" test fixture that would build an enclave with a large heap and then have reclaimer tests that exercise it by being tests that are specific to this "reclaimer fixture"?
Quoted 4 lines
Quoted 1 line
Reinette
Why add that complexity?
With this change every test is turned into a pseudo reclaimer test without there being any explicit testing (with pass/fail criteria) of reclaimer behavior. This is an expensive addition and reduces the scenarios that the tests can exercise.
Quoted 6 lines
Quoted 1 line
Reinette
There is consistent known behaviour how reclaimer and also the page fault are exercised for each test. I think that is what matters most right now that the basic behaviour of both the page reclaimer and page fault handler gets exercised.
I believe the basic behavior of page fault handler is currently exercised in each test, this is required.
Quoted 4 lines
I don't understand the real-world gain of doing something factors more complex than necessary at a particular point of time, when you don't really need to hang yourself into it forever.
Your argument about "hang yourself into it forever" can go both ways - why should all tests now unnecessarily consume the entire EPC forever?
If I understand correctly adding a separate reclaimer test is not complex but would require refactoring code.
Quoted 4 lines
This patch does increase the coverage in a deterministic manner to the code paths that were not previously exercised, i.e. we know the code paths, and could even calculate the exact number of times that they are triggered. And without doing anything obscure. That's what matters to me.
On the contrary this is indeed obfuscating the SGX tests: if an issue shows up in the reclaimer then all tests would fail. If there is a unique reclaimer test then that would help point to where the issue may be.
Reinette
Jarkko Sakkinen
maintainerRe: [PATCH 4/4] selftests/sgx: Trigger the reclaimer and #PF handler
Quoted 62 lines · Reinette Chatre
Hi Jarkko,
Quoted 59 lines · Jarkko Sakkinen
Quoted 57 lines · Reinette Chatre
Hi Jarkko,
Quoted 43 lines · Jarkko Sakkinen
Quoted 42 lines · Reinette Chatre
Hi Jarkko,
Quoted 27 lines · Jarkko Sakkinen
Quoted 26 lines · Reinette Chatre
Hi Jarkko,
Quoted 7 lines · Jarkko Sakkinen
Create a heap for the test enclave, which has the same size as all available Enclave Page Cache (EPC) pages in the system. This will guarantee that all test_encl.elf pages *and* SGX Enclave Control Structure (SECS) have been swapped out by the page reclaimer during the load time. Actually, this adds a bit more stress than that since part of the EPC gets reserved for the Version Array (VA) pages.
Quoted 7 lines
For each test, the page fault handler gets triggered in two occasions:
- When
SGX_IOC_ENCLAVE_INITis performed, SECS gets swapped in by the page fault handler.
- During the execution, each page that is referenced gets swapped in by the page fault handler.If I understand this correctly, all EPC pages are now being consumed during fixture setup and thus every SGX test, no matter how big or small, now becomes a stress test of the reclaimer instead of there being a unique reclaimer test. Since an enclave is set up and torn down for every test this seems like a significant addition. It also seems like this would impact future tests of dynamic page addition where not all scenarios could be tested with all EPC pages already consumed.
Quoted 5 lines
Quoted 1 line
Reinette
Re-initializing the test enclave is mandatory thing to do for all tests because it has an internals state.
Right, but not all tests require the same enclave. In kselftest terminology I think you are attempting to force all tests to depend on the same test fixture. Is it not possible to have a separate "reclaimer" test fixture that would build an enclave with a large heap and then have reclaimer tests that exercise it by being tests that are specific to this "reclaimer fixture"?
Quoted 4 lines
Quoted 1 line
Reinette
Why add that complexity?
With this change every test is turned into a pseudo reclaimer test without there being any explicit testing (with pass/fail criteria) of reclaimer behavior. This is an expensive addition and reduces the scenarios that the tests can exercise.
Reinette
Quoted 7 lines
Quoted 4 lines
There is consistent known behaviour how reclaimer and also the page fault are exercised for each test. I think that is what matters most right now that the basic behaviour of both the page reclaimer and page fault handler gets exercised.
I believe the basic behavior of page fault handler is currently exercised in each test, this is required.
This not true. The current test does not exercise ELDU code path.
Quoted 11 lines
Quoted 5 lines
I don't understand the real-world gain of doing something factors more complex than necessary at a particular point of time, when you don't really need to hang yourself into it forever.
Your argument about "hang yourself into it forever" can go both ways - why should all tests now unnecessarily consume the entire EPC forever?
If I understand correctly adding a separate reclaimer test is not complex but would require refactoring code.
What does it matter anyway if code nees to be refactored?
Quoted 8 lines
Quoted 4 lines
This patch does increase the coverage in a deterministic manner to the code paths that were not previously exercised, i.e. we know the code paths, and could even calculate the exact number of times that they are triggered. And without doing anything obscure. That's what matters to me.
On the contrary this is indeed obfuscating the SGX tests: if an issue shows up in the reclaimer then all tests would fail. If there is a unique reclaimer test then that would help point to where the issue may be.
I tend to disagree this. I'll add a separate reclaimer test if I need to test something that this does not scale. It's an iterative process.
/Jarkko
Re: [PATCH 4/4] selftests/sgx: Trigger the reclaimer and #PF handler
In reply to Jarkko SakkinenQuoted 4 lines · Jarkko Sakkinen
There is consistent known behaviour how reclaimer and also the page fault are exercised for each test. I think that is what matters most right now that the basic behaviour of both the page reclaimer and page fault handler gets exercised.
There's also a lot of value to ensuring that tests can run _quickly_. If you have a test that fails one out of a million executions, it's a lot easier find and debug if it takes 1 ms versus 10 seconds.
In other words, I think I'd prefer if we run two enclaves in each execution of the selftest. One can be as small as possible. The other can be the reclaim-triggering one.
That's good both for test coverage, and it makes it a *bit* more straightforward to hack out the reclaim test if you need things to run faster.
The pkeys selftest isn't a bad example here either. It has a couple of different "malloc()" options: THP, hugetlbfs, small-page mmap(), and a bunch of tests it runs on each type. As we add more SGX tests, we might end up with "do reclaim" just being an option we pass.
Jarkko Sakkinen
maintainerRe: [PATCH 4/4] selftests/sgx: Trigger the reclaimer and #PF handler
Quoted 14 lines · Dave Hansen
Quoted 4 lines · Jarkko Sakkinen
There is consistent known behaviour how reclaimer and also the page fault are exercised for each test. I think that is what matters most right now that the basic behaviour of both the page reclaimer and page fault handler gets exercised.
There's also a lot of value to ensuring that tests can run _quickly_. If you have a test that fails one out of a million executions, it's a lot easier find and debug if it takes 1 ms versus 10 seconds.
In other words, I think I'd prefer if we run two enclaves in each execution of the selftest. One can be as small as possible. The other can be the reclaim-triggering one.
Quoted 8 lines
That's good both for test coverage, and it makes it a *bit* more straightforward to hack out the reclaim test if you need things to run faster.
The pkeys selftest isn't a bad example here either. It has a couple of different "
malloc()" options: THP, hugetlbfs, small-pagemmap(), and a bunch of tests it runs on each type. As we add more SGX tests, we might end up with "do reclaim" just being an option we pass.
Even with large EPC's, the current test runs quite fast, because heap is left unmeasured. It's the EEXTEND operations that would cause a major slow-down.
I would go only to something "more complex" when the current test hits the roof. I don't like to make code more complicated, when that does not happen.
When there's no compatibility requirements, it's not hard to refactor it later on.
/Jarkko